Dear SPM experts:
For a two condition single-event experiment (ISI=8.2s) with events of
type A or B, where B is a slight variation of A thereby taking advantage
of the principle of cognitive subtraction...
We have analyzed the data first by looking at the effect of one
covariate of interest at a time (i.e., [1,0] or [0,1]). Having not seen
any of the hypothesized activation regions, we further
asked whether the magnitude of one effect was significantly greater than
that of the other with two additional contrasts [1 -1] and [-1 1].
These latter two contrasts give considerably
different SPMs compared to the ones that isolate a single effect. This
combination of results suggest to us that the [1 -1] and [-1 1]
contrasts may be more sensitive to relevant task
associated changes than a more typical contrast settings used for
event-related experiments? Does this make statistical sense, and are
there any published methodological references which document and further
describe such a phenomenon?
Any assistance you can offer will be appreciated.
Thanks, Michael
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|